Simulation based Planning, Implementing and Interpretation of Results of Foam Treatment to Reduce GOR

Abstract

Horizontal oil wells drilled into the oil rim in the Caspian region are long and quite productive, but were producing much more gas than expected. The Operator did a few studies to understand the source of gas: gas was coming with oil, i.e. plugging these intervals would result in the oil shortage. The Operator paid attention to the experience of the Norwegian Continental Shelf (NCS), where foam treatments were implemented for this purpose. The Operator had no information from the reservoir, but decided reasonable to ask PETEC to build a reservoir model based on open data from the NCS and expert opinion. Simultaneously the Operator ordered lab experiments to measure gas/oil mobility vs surfactant concentration from another company. The model was built in CMG STARS, which allows using either mechanistic or empirical approach to evaluate the effect of the chemical treatment in a well. The empirical approach was chosen due to very limited amount of information. At the later stage the lab data arrived allowing to update the model with the actual reservoir related data. During 2012-2013 the Operator obtained more lab results and adjusted the initial plan. At the end of 2013 the treatment was implemented on a well. The effect on GOR was less than expected: the gas production dropped, but restored rapidly, although not reaching the initial. The Operator asked PETEC to update the reservoir model and interpret the results of the treatment using the data obtained and accumulated within 2012-2013. This included: (i) lab data on adsorption, desorption, chemisorption, compatibility of water; foam stability and gas/oil mobility vs surfactant concentration; (ii) actual well performance, pressures and temperatures; (iii) actual treatment procedure record vs plan. The model was updated with this data, significantly increasing its accuracy. However, that wasn’t enough to get accurate prognosis. So the multi-scenario runs were demanded again to match observed values to the model. There were a few different combinations of variable parameters matching the model. They were split into 2-3 “situations” in the reservoir and well. Based on this the program for lab experiments was handed over to the Operator. Executing this program should allow the Operator to understand, which “situation” happened in reality. The paper explains the approaches implemented in this work and gives the procedure of building the empirical model to evaluate the effect of surfactant (foam) treatment in oil wells. It’s also important, that the work is not just simulation, but the real results were obtained and analyzed.